Satisfiability Solvers

1. Satisfiability Solvers Part 2: Stochastic Solvers 600.325/425 Declarative Methods - J. Eisner

2. Structured vs. Random Problems • So far, we’ve been dealing with SAT problems that encode other problems • Most not as hard as # of variables & clauses suggests • Small crossword grid + medium-sized dictionary may turn into a big formula … but still a small puzzle at some level • Unit propagation does a lot of work for you • Clause learning picks up on the structure of the encoding • But some random SAT problems really are hard! • zChaff’s tricks don’t work so well here 600.325/425 Declarative Methods - J. Eisner

3. Random 3-SAT • sample uniformly from space of all possible 3-clauses • n variables, l clauses • Which are the hard instances? • around l/n = 4.26 600.325/425 Declarative Methods - J. Eisner slide thanks to Henry Kautz (modified)

4. Complexity peak is very stable … • across problem sizes • across solver types • systematic (last lecture) • stochastic (this lecture) The magic ratio 4.26 600.325/425 Declarative Methods - J. Eisner slide thanks to Henry Kautz (modified)

5. Why 4.26? • Complexity peak coincides with solubility transition • l/n < 4.3 problems under-constrained and SAT • l/n > 4.3 problems over-constrained and UNSAT • l/n=4.3, problems on “knife-edge” between SAT and UNSAT 600.325/425 Declarative Methods - J. Eisner slide thanks to Henry Kautz (modified)

6. That’s called a “phase transition” • Problems < 32˚F are like ice; > 32˚F are like water • Similar “phase transitions” for other NP-hard problems • job shop scheduling • traveling salesperson (instances from TSPlib) • exam timetables (instances from Edinburgh) • Boolean circuit synthesis • Latin squares (alias sports scheduling) • Hot research topic: • predict hardness of a given instance, & use hardness to control search strategy (Horvitz, Kautz, Ruan 2001-3) 600.325/425 Declarative Methods - J. Eisner slide thanks to Henry Kautz (modified)

7. Methods in today’s lecture … • Can handle “big” random SAT problems • Can go up about 10x bigger than systematic solvers!  • Rather smaller than structured problems (espec. if ratio  4.26) • Also handle big structured SAT problems • But lose here to best systematic solvers  • Try hard to find a good solution • Very useful for approximating MAX-SAT • Not intended to find all solutions • Not intended to show that there are no solutions (UNSAT) 600.325/425 Declarative Methods - J. Eisner

8. GSAT vs. DP on Hard Random Instances today (not today’s best algorithm) yesterday (not yesterday’s best algorithm) 600.325/425 Declarative Methods - J. Eisner slide thanks to Russ Greiner and Dekang Lin

9. How’m I doin’? Repeat until satisfied Local search for SAT • Make a guess (smart or dumb) about values of the variables • Try flipping a variable to make things better • Algorithms differ on which variable to flip 600.325/425 Declarative Methods - J. Eisner

10. Make a guess (smart or dumb) about values of the variables • Try flipping a variable that makes things better How’m I doin’? Repeat until satisfied Local search for SAT • Flip a randomly chosen variable? • No, blundering around blindly takes exponential time • Ought to pick a variable that improves … what? • Increase the # of satisfied clauses as much as possible • Break ties randomly • Note: Flipping a var will repair some clauses and break others • This is the “GSAT” (Greedy SAT) algorithm (almost) 600.325/425 Declarative Methods - J. Eisner

11. Make a guess (smart or dumb) about values of the variables • Try flipping a variable that makes things better How’m I doin’? Repeat until satisfied Local search for SAT • Flip most-improving variable. Guaranteed to work? • What if first guess is A=1, B=1, C=1? • 2 clauses satisfied • Flip A to 0  3 clauses satisfied • Flip B to 0  all 4 clauses satisfied (pick this!) • Flip C to 0  3 clauses satisfied A v C A v B ~C v ~B ~B v ~A 600.325/425 Declarative Methods - J. Eisner example thanks to Monaldo Mastrolilli and Luca Maria Gambardella

12. Make a guess (smart or dumb) about values of the variables • Try flipping a variable that makes things better How’m I doin’? Repeat until satisfied Local search for SAT • Flip most-improving variable. Guaranteed to work? • But what if first guess is A=0, B=1, C=1? • 3 clauses satisfied • Flip A to 1  3 clauses satisfied • Flip B to 0  3 clauses satisfied • Flip C to 0  3 clauses satisfied • Pick one anyway … (picking A wins on next step) A v C A v B ~C v ~B ~B v ~A 600.325/425 Declarative Methods - J. Eisner example thanks to Monaldo Mastrolilli and Luca Maria Gambardella

13. Make a guess (smart or dumb) about values of the variables • Try flipping a variable that makes things better How’m I doin’? Repeat until satisfied Local search for SAT • Flip most-improving variable. Guaranteed to work? • Yes for 2-SAT: probability  1 within O(n2) steps • No in general: can & usually does get locally stuck • Therefore, GSAT just restarts periodically • New random initial guess Believe it or not, GSAT outperformed everything else when it was invented in 1992. 600.325/425 Declarative Methods - J. Eisner example thanks to Monaldo Mastrolilli and Luca Maria Gambardella

14. Problems with Hill Climbing 600.325/425 Declarative Methods - J. Eisner slide thanks to Russ Greiner and Dekang Lin

15. Problems with Hill Climbing • Local Optima (foothills): No neighbor is better, but not at global optimum. • (Maze: may have to move AWAY from goal to find (best) solution) • Plateaus: All neighbors look the same. • (15-puzzle: perhaps no action will change # of tiles out of place) • Ridge: going up only in a narrow direction. • Suppose no change going South, or going East, but big win going SE: have to flip 2 vars at once • Ignorance of the peak: Am I done? 600.325/425 Declarative Methods - J. Eisner slide thanks to Russ Greiner and Dekang Lin

16. How to escape local optima? • Restart every so often • Don’t be so greedy (more randomness) • Walk algorithms: With some probability, flip a randomly chosen variable instead of a best-improving variable • WalkSAT algorithms: Confine selection to variables in a single, randomly chosen unsatisfied clause (also faster!) • Simulated annealing (general technique): Probability of flipping is related to how much it helps/hurts • Force the algorithm to move away from current solution • Tabu algs: Refuse to flip back a var flipped in past t steps • Novelty algs for WalkSAT: With some probability, refuse to flip back the most recently flipped var in a clause • Dynamic local search algorithms: Gradually increase weight of unsatisfied clauses 600.325/425 Declarative Methods - J. Eisner

17. How to escape local optima? Lots of algorithms have been tried … Simulated annealing, genetic or evolutionary algorithms, GSAT, GWSAT, GSAT/Tabu, HSAT, HWSAT, WalkSAT/SKC, WalkSAT/Tabu, Novelty, Novelty+, R-Novelty(+), Adaptive Novelty(+), … 600.325/425 Declarative Methods - J. Eisner

18. How do we generalize to MAX-SAT? (the “novelty” part) Adaptive Novelty+ (current winner) • with probability 1% (the “+” part) • Choose randomly among all variables that appear in at least one unsatisfied clause • else be greedier (other 99%) • Randomly choose an unsatisfied clause C • Flipping any variable in C will at least fix C • Choose the most-improving variable in C … except … • with probability p, ignore most recently flipped variable in C • The “adaptive” part: • If we improved # of satisfied clauses, decrease p slightly • If we haven’t gotten an improvement for a while, increase p • If we’ve been searching too long, restart the whole algorithm 600.325/425 Declarative Methods - J. Eisner

19. WalkSAT/SKC (first good local search algorithm for SAT, very influential) • Choose an unsatisfied clause C • Flipping any variable in C will at least fix C • Compute a “break score” for each var in C • Flipping it would break how many other clauses? • If C has any vars with break score 0 • Pick at random among the vars with break score 0 • else with probability p • Pick at random among the vars with min break score • else • Pick at random among all vars in C 600.325/425 Declarative Methods - J. Eisner

20. Simulated Annealing (popular general local search technique – often very effective, rather slow) • Pick a variable at random If flipping it improves assignment: do it. Else flip anyway with probability p = e-/T where •  = damage to score • What is p for =0? For large ? • T = “temperature” • What is p as T tends to infinity? • Higher T = more random, non-greedy exploration • As we run, decrease T from high temperature to near 0. 600.325/425 Declarative Methods - J. Eisner slide thanks to Russ Greiner and Dekang Lin (modified)

21. Evolutionary algorithms (another popular general technique) • Many local searches at once • Consider 20 random initial assignments • Each one gets to try flipping 3 different variables(“reproduction with mutation”) • Now we have 60 assignments • Keep the best 20 (“natural selection”), and continue 600.325/425 Declarative Methods - J. Eisner

22. 1 0 1 0 1 1 1 Parent 1 1 1 0 0 0 1 1 Parent 2 1 0 10 0 1 1 Child 1 Mutation Child 2 1 1 00 1 1 0 Sexual reproduction (another popular general technique – at least for evolutionary algorithms) • Derive each new assignment by somehow combining two old assignments, not just modifying one (“sexual reproduction” or “crossover”) Good idea? 600.325/425 Declarative Methods - J. Eisner slide thanks to Russ Greiner and Dekang Lin (modified)